diff mbox series

[11/16] block: add optimised version bio_set_dev()

Message ID 3c908cb74959c631995341111a7ce116487da5c5.1634676157.git.asml.silence@gmail.com (mailing list archive)
State New, archived
Headers show
Series block optimisation round | expand

Commit Message

Pavel Begunkov Oct. 19, 2021, 9:24 p.m. UTC
If a bio was just allocated its flags should be zero and there is no
need to clear them. Add __bio_set_dev(), which is faster and doesn't
care about clering flags.

Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
---
 block/fops.c        |  4 ++--
 include/linux/bio.h | 10 ++++++++--
 2 files changed, 10 insertions(+), 4 deletions(-)

Comments

Chaitanya Kulkarni Oct. 19, 2021, 10:36 p.m. UTC | #1
On 10/19/21 2:24 PM, Pavel Begunkov wrote:
> External email: Use caution opening links or attachments
> 
> 
> If a bio was just allocated its flags should be zero and there is no
> need to clear them. Add __bio_set_dev(), which is faster and doesn't
> care about clering flags.
> 
> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>

Looks good.

Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Christoph Hellwig Oct. 20, 2021, 6:20 a.m. UTC | #2
On Tue, Oct 19, 2021 at 10:24:20PM +0100, Pavel Begunkov wrote:
> If a bio was just allocated its flags should be zero and there is no
> need to clear them. Add __bio_set_dev(), which is faster and doesn't
> care about clering flags.

This is entirely the wrong way around.  Please add a new bio_reset_dev
for the no more than about two hand full callers that actually had
a device set and just optimize bio_set_dev.
Pavel Begunkov Oct. 20, 2021, 12:29 p.m. UTC | #3
On 10/20/21 07:20, Christoph Hellwig wrote:
> On Tue, Oct 19, 2021 at 10:24:20PM +0100, Pavel Begunkov wrote:
>> If a bio was just allocated its flags should be zero and there is no
>> need to clear them. Add __bio_set_dev(), which is faster and doesn't
>> care about clering flags.
> 
> This is entirely the wrong way around.  Please add a new bio_reset_dev
> for the no more than about two hand full callers that actually had
> a device set and just optimize bio_set_dev.

makes sense, thanks
diff mbox series

Patch

diff --git a/block/fops.c b/block/fops.c
index 8e3790faafb8..7cf98db0595a 100644
--- a/block/fops.c
+++ b/block/fops.c
@@ -75,7 +75,7 @@  static ssize_t __blkdev_direct_IO_simple(struct kiocb *iocb,
 	}
 
 	bio_init(&bio, vecs, nr_pages);
-	bio_set_dev(&bio, bdev);
+	__bio_set_dev(&bio, bdev);
 	bio.bi_iter.bi_sector = pos >> 9;
 	bio.bi_write_hint = iocb->ki_hint;
 	bio.bi_private = current;
@@ -224,7 +224,7 @@  static ssize_t __blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter,
 		blk_start_plug(&plug);
 
 	for (;;) {
-		bio_set_dev(bio, bdev);
+		__bio_set_dev(bio, bdev);
 		bio->bi_iter.bi_sector = pos >> 9;
 		bio->bi_write_hint = iocb->ki_hint;
 		bio->bi_private = dio;
diff --git a/include/linux/bio.h b/include/linux/bio.h
index c88700d1bdc3..0ab4fa2c89c3 100644
--- a/include/linux/bio.h
+++ b/include/linux/bio.h
@@ -453,13 +453,19 @@  static inline void bio_clone_blkg_association(struct bio *dst,
 					      struct bio *src) { }
 #endif	/* CONFIG_BLK_CGROUP */
 
+/* can be used only with freshly allocated bios */
+static inline void __bio_set_dev(struct bio *bio, struct block_device *bdev)
+{
+	bio->bi_bdev = bdev;
+	bio_associate_blkg(bio);
+}
+
 static inline void bio_set_dev(struct bio *bio, struct block_device *bdev)
 {
 	bio_clear_flag(bio, BIO_REMAPPED);
 	if (bio->bi_bdev != bdev)
 		bio_clear_flag(bio, BIO_THROTTLED);
-	bio->bi_bdev = bdev;
-	bio_associate_blkg(bio);
+	__bio_set_dev(bio, bdev);
 }
 
 static inline void bio_copy_dev(struct bio *dst, struct bio *src)